amdgpu_asic_flush_hdp(adev, ring);
}
+ if (need_ctx_switch)
+ status |= AMDGPU_HAVE_CTX_SWITCH;
+
skip_preamble = ring->current_ctx == fence_ctx;
if (job && ring->funcs->emit_cntxcntl) {
- if (need_ctx_switch)
- status |= AMDGPU_HAVE_CTX_SWITCH;
status |= job->preamble_status;
-
amdgpu_ring_emit_cntxcntl(ring, status);
}
!amdgpu_sriov_vf(adev)) /* for SRIOV preemption, Preamble CE ib must be inserted anyway */
continue;
- amdgpu_ring_emit_ib(ring, job, ib, need_ctx_switch);
- need_ctx_switch = false;
+ amdgpu_ring_emit_ib(ring, job, ib, status);
+ status &= ~AMDGPU_HAVE_CTX_SWITCH;
}
if (ring->funcs->emit_tmz)
void (*emit_ib)(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch);
+ uint32_t flags);
void (*emit_fence)(struct amdgpu_ring *ring, uint64_t addr,
uint64_t seq, unsigned flags);
void (*emit_pipeline_sync)(struct amdgpu_ring *ring);
#define amdgpu_ring_get_rptr(r) (r)->funcs->get_rptr((r))
#define amdgpu_ring_get_wptr(r) (r)->funcs->get_wptr((r))
#define amdgpu_ring_set_wptr(r) (r)->funcs->set_wptr((r))
-#define amdgpu_ring_emit_ib(r, job, ib, c) ((r)->funcs->emit_ib((r), (job), (ib), (c)))
+#define amdgpu_ring_emit_ib(r, job, ib, flags) ((r)->funcs->emit_ib((r), (job), (ib), (flags)))
#define amdgpu_ring_emit_pipeline_sync(r) (r)->funcs->emit_pipeline_sync((r))
#define amdgpu_ring_emit_vm_flush(r, vmid, addr) (r)->funcs->emit_vm_flush((r), (vmid), (addr))
#define amdgpu_ring_emit_fence(r, addr, seq, flags) (r)->funcs->emit_fence((r), (addr), (seq), (flags))
void amdgpu_vce_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
amdgpu_ring_write(ring, VCE_CMD_IB);
amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx);
int amdgpu_vce_ring_parse_cs_vm(struct amdgpu_cs_parser *p, uint32_t ib_idx);
void amdgpu_vce_ring_emit_ib(struct amdgpu_ring *ring, struct amdgpu_job *job,
- struct amdgpu_ib *ib, bool ctx_switch);
+ struct amdgpu_ib *ib, uint32_t flags);
void amdgpu_vce_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
unsigned flags);
int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring);
static void cik_sdma_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 extra_bits = vmid & 0xf;
static void gfx_v6_0_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 header, control = 0;
/* insert SWITCH_BUFFER packet before first IB in the ring frame */
- if (ctx_switch) {
+ if (flags & AMDGPU_HAVE_CTX_SWITCH) {
amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
amdgpu_ring_write(ring, 0);
}
static void gfx_v7_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 header, control = 0;
/* insert SWITCH_BUFFER packet before first IB in the ring frame */
- if (ctx_switch) {
+ if (flags & AMDGPU_HAVE_CTX_SWITCH) {
amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
amdgpu_ring_write(ring, 0);
}
static void gfx_v7_0_ring_emit_ib_compute(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vmid << 24);
static void gfx_v8_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 header, control = 0;
static void gfx_v8_0_ring_emit_ib_compute(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vmid << 24);
static void gfx_v9_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 header, control = 0;
static void gfx_v9_0_ring_emit_ib_compute(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vmid << 24);
static void sdma_v2_4_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void sdma_v3_0_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void sdma_v4_0_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void si_dma_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
/* The indirect buffer packet must end on an 8 DW boundary in the DMA ring.
static void uvd_v4_2_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
amdgpu_ring_write(ring, PACKET0(mmUVD_RBC_IB_BASE, 0));
amdgpu_ring_write(ring, ib->gpu_addr);
static void uvd_v5_0_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
amdgpu_ring_write(ring, PACKET0(mmUVD_LMI_RBC_IB_64BIT_BAR_LOW, 0));
amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
static void uvd_v6_0_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void uvd_v6_0_enc_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void uvd_v7_0_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
struct amdgpu_device *adev = ring->adev;
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void uvd_v7_0_enc_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void vce_v3_0_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
#endif
static void vce_v4_0_ring_emit_ib(struct amdgpu_ring *ring, struct amdgpu_job *job,
- struct amdgpu_ib *ib, bool ctx_switch)
+ struct amdgpu_ib *ib, uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void vcn_v1_0_dec_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
struct amdgpu_device *adev = ring->adev;
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void vcn_v1_0_enc_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
static void vcn_v1_0_jpeg_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
- bool ctx_switch)
+ uint32_t flags)
{
struct amdgpu_device *adev = ring->adev;
unsigned vmid = AMDGPU_JOB_GET_VMID(job);